Torn tendons and ligaments often require surgical repair to their bony insertions. A large percentage of these repairs have poor outcomes; for example, up to 94% of surgical rotator cuff repairs fail. At the root of these failures is the fundamental challenge of attaching two materials, tendon and bone, with vastly different mechanical properties. The natural tendon-to-bone insertion involves a number of mechanisms that create a strong and tough attachment. Unfortunately, this tissue degrades with age, and is not regenerated in healing. Our overall goal is to develop a multiscale model of the tendon-to-bone insertion that will lead to (1) tissue toughness metrics that can guide clinical decisions for elderly patients, and (2) foundations for future tissue- engineered surgical grafts. Based on our previous work, we hypothesize that toughening and strengthening mechanisms exist across several length scales, and that these are most pronounced in a the compliant region of tissue between tendon and bone that does not regrow in the healing setting. We will characterize the stiffening, strengthening, and toughening mechanisms that contribute to this resilience across scales in natural and pathologic tendon-to-bone insertions as a function of age. The work involves three aims: (1) At the nanoscale, elemental spatial maps will be acquired using transmission electron microscopy electron energy loss spectroscopy to determine mineral and collagen distributions across the insertion. Individual mineralized collagen fibrils will be mechanically tested; we have recently performed such tests on mammalian collagen fibrils. In silico experiments will identify and quantify deformation mechanisms underlying the toughness of mineralized collagen fibrils. (2) At the microscale, synchrotron X-ray diffraction, Raman spectroscopy, and polarized light microscopy will be used to determine the distributions of mineral content and collagen orientation. Mechanics of the tendon-to-bone insertion will be examined with micrometer resolution using a confocal microscope-mounted testing frame. In silico, nonlinear homogenization methods will be used to incorporate mineralized collagen fiber mechanics from Aim 1 into constitutive models of connected networks of mineralized and cross-linked collagen fibers. (3) At the millimeter scale, the 3D inter-digitation geometry of tendon and bone will be determined using phase contrast X-ray computed tomography and the mechanics of the tendon-to-bone insertion will be determined using tissue level tensile tests. In silico experiments combining tendon-to-bone geometry with microscale tissue models will produce hypotheses of mechanisms underlying tendon-to-bone insertion toughness. Mechanical fields will be passed down hierarchical model levels to evaluate collagen fibril response to predicted physiologic and pathologic tendon-to-bone insertion loading. Together, these models and data form the foundation of future tissue engineering efforts and efforts to identify clinically useful metrics of tendon-to-bone tissue health.